Kernel-Ridge Regression-Based Quality Measure and Enhancement of Three-Dimensional-Synthesized Images

Jakhetiya, Vinit; Gu, Ke; Jaiswal, Sunil Prasad; Singhal, Trisha; Xia, Zhifang

doi:10.1109/tie.2020.2965469

Cited by 19 publications

(8 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The scores are mapped to the ground truth of each view pair using the non-linearity equation as given in [3]. [8] NR IQA 0.6881 0.6261 3 SC-IQA [14] FR IQA 0.6620 0.5960 4 DSCB [9] NR IQA 0.6030 0.5571 5 LOGS [15] FR IQA 0.6280 0.6160 6 MP-PSNR [16] FR IQA 0.6190 0.5809 7 MW-PSNR [17] FR IQA 0.5389 0.4875 8 Wang's [18] NR IQA 0.4338 0.4254 9 APT [3] NR IQA 0.4329 0.4164 10 Jakhetiya [19] NR IQA 0.2715 0.2352 11 OMIQA [20] NR IQA 0.2705 0.2331 12 NIQSV+ [21] NR IQA 0.2324 0.1545…”

Section: Experiments Results Analysismentioning

confidence: 99%

“…Also, these results will motivate future researchers to use cosine similarity instead of MSE when distortions are not uniformly distributed across the images. Further,in order to show that the proposed algorithm is performing better than the existing ones, we compared the it with the recently proposed five FR (SSPD [10], SC-IQA [14], LOGS [15], MP-PSNR [16], MW-PSNR [17]) and seven NR quality assessment algorithms (Yan's [8], DSCB [9], Wang's [18], APT [3], Jakhetiya's [19], OMIQA [20], NIQSV+ [21]). As depicted in Table 3, the proposed method outperforms the contemporary IQA metrics developed for DIBR synthesized views.…”

Section: Experiments Results Analysismentioning

confidence: 99%

See 1 more Smart Citation

Perceptual Quality Assessment of DIBR Synthesized Views Using Saliency Based Deep Features

Chaudhary

Mazumder²,

Mumtaz

et al. 2021

2021 IEEE International Conference on Image Processing (ICIP)

Self Cite

View full text Add to dashboard Cite

In recent years, Depth-Image-Based-Rendering (DIBR) synthesized views have gained popularity due to their numerous visual media applications. Consequently, the research in their quality assessment (QA) has also gained momentum. In this work, we propose an efficient metric to estimate the perceptual quality of DIBR synthesized views via the extraction of Deep-features. These Deep-features are extracted from a pretrained CNN model. Generally, in DIBR synthesized views, geometric distortions arise near the objects due to occlusion, and the human visual system is quite sensitive towards these objects. On the other end, saliency maps are efficiently able to highlight perceptually important objects. With this intuition, instead of extracting deep features directly from DIBR synthesized views, we obtain the refined feature vector from their corresponding saliency maps. Also, most of the pixels with geometric distortions have a nearly similar impact on the perceptual quality of 3D synthesized views. Considering this, we propose to fuse the feature maps using the cosine similarity measure based upon the deviation of one feature vector from another. It may also be emphasized that no training is performed in the proposed algorithm, and all the features are extracted from the pre-trained vanilla VGG-16 architecture. The proposed metric, when applied to the standard database, results in PLCC of 0.762 and SRCC equal to 0.7513, which is better than the existing state-of-the-art QA metrics.

show abstract

Section: Experiments Results Analysismentioning

confidence: 99%

Section: Experiments Results Analysismentioning

confidence: 99%

Perceptual Quality Assessment of DIBR Synthesized Views Using Saliency Based Deep Features

Chaudhary

Mazumder²,

Mumtaz

et al. 2021

2021 IEEE International Conference on Image Processing (ICIP)

Self Cite

View full text Add to dashboard Cite

show abstract

“…With this concern, some researchers have proposed IQA metrics targeting 3D synthesized images. These methods are mainly divided into two categories, full-reference (FR) [ 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 ] and no-reference (NR) [ 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 ].…”

Section: Introductionmentioning

confidence: 99%

“…Jakhetiya et al counted outliers by a three sigma rule-based robust outlyingness ratio (OUT) to evaluate the quality of synthesized images [ 26 ]. Recently, Jakhetiya et al further proposed a kernel-ridge-regression-based predictor for synthesized IQA, which detected the complete distortion surface with geometric distortions and estimated corresponding quality scores [ 27 ]. The NSS-based methods above are time consuming and basically designed for severe geometric distortions.…”

Section: Introductionmentioning

confidence: 99%

No-Reference Quality Assessment for 3D Synthesized Images Based on Visual-Entropy-Guided Multi-Layer Features Analysis

Jin

Peng

Zou

et al. 2021

Entropy

View full text Add to dashboard Cite

Multiview video plus depth is one of the mainstream representations of 3D scenes in emerging free viewpoint video, which generates virtual 3D synthesized images through a depth-image-based-rendering (DIBR) technique. However, the inaccuracy of depth maps and imperfect DIBR techniques result in different geometric distortions that seriously deteriorate the users’ visual perception. An effective 3D synthesized image quality assessment (IQA) metric can simulate human visual perception and determine the application feasibility of the synthesized content. In this paper, a no-reference IQA metric based on visual-entropy-guided multi-layer features analysis for 3D synthesized images is proposed. According to the energy entropy, the geometric distortions are divided into two visual attention layers, namely, bottom-up layer and top-down layer. The feature of salient distortion is measured by regional proportion plus transition threshold on a bottom-up layer. In parallel, the key distribution regions of insignificant geometric distortion are extracted by a relative total variation model, and the features of these distortions are measured by the interaction of decentralized attention and concentrated attention on top-down layers. By integrating the features of both bottom-up and top-down layers, a more visually perceptive quality evaluation model is built. Experimental results show that the proposed method is superior to the state-of-the-art in assessing the quality of 3D synthesized images.

show abstract

“…In the latter, the number of parameters to be determined is not fixed, but depends on the number of available data-points. This non-parametric approach to building models is popular in many disciplines, usually in the form of Gaussian processes (GPs)-whose means are weighted sums of kernels-or plain radial basis functions (RBFs) [1]- [3]. Among systems and control researchers, kernel methods have also been studied with the aim of adapting and improving existing tools (see [4]- [6] for recent reviews).…”

Section: Introductionmentioning

confidence: 99%

Deterministic error bounds for kernel-based learning techniques under bounded noise

Maddalena¹,

Scharnhorst²,

Jones³

2020

Preprint

View full text Add to dashboard Cite

We consider the problem of reconstructing a function from a finite set of noise-corrupted samples. Two kernel algorithms are analyzed, namely kernel ridge regression and εsupport vector regression. By assuming the ground-truth function belongs to the reproducing kernel Hilbert space of the chosen kernel, and the measurement noise affecting the dataset is bounded, we adopt an approximation theory viewpoint to establish deterministic error bounds for the two models. Finally, we discuss their connection with Gaussian processes and two numerical examples are provided. In establishing our inequalities, we hope to help bring the fields of non-parametric kernel learning and robust control closer to each other.

show abstract

Kernel-Ridge Regression-Based Quality Measure and Enhancement of Three-Dimensional-Synthesized Images

Cited by 19 publications

References 40 publications

Perceptual Quality Assessment of DIBR Synthesized Views Using Saliency Based Deep Features

Perceptual Quality Assessment of DIBR Synthesized Views Using Saliency Based Deep Features

No-Reference Quality Assessment for 3D Synthesized Images Based on Visual-Entropy-Guided Multi-Layer Features Analysis

Deterministic error bounds for kernel-based learning techniques under bounded noise

Contact Info

Product

Resources

About